By far, there are already some online customer service products on the market. These products mostly provide an online indexing and interactive Q&A ability by converting the existing knowledge into an online knowledge repository, thus improving the work efficiency and automation of the customer service.
However, these existing online customer service products mainly focus on the indexing of knowledge and the voice interaction to expand applications, and do not drill down to the real-time status layer of the specific devices. They cannot precisely locate problems and still belong to the empirical system category, having major blind zones and uncertainty. Moreover, the existing technical architecture also does not have the ability to automatically explore new knowledge paths and learn new knowledge by themselves.